I. Ruler axiom

Statement

Let $\ell$ be a line. There exists a bijection

$$f : \ell \;\longleftrightarrow\; \mathbb{R}$$

such that for any two points $A$, $B$ on $\ell$, the distance

$$|AB| \;=\; |\,f(A) - f(B)\,|.$$

Each such $f$ is called a coordinate system on $\ell$.

Intuition

Glue the entire real line onto the line $\ell$: every point on the line is labeled with a real number, and the distance between two points is the absolute value of the difference of their coordinates.

This single axiom does two things at once — it tells us what a line "looks like" (isomorphic to the reals) and how to "measure" on it (distance comes from real-number subtraction). Those vague notions of "length" in Euclidean geometry, from this moment on, acquire a precise algebraic expression.

Ruler freedom: $f$ is not unique

The axiom only promises the existence of such an $f$ — it does not say "there is only one." To get to the bottom of $f$'s degrees of freedom, three things need to be made clear first.

1. $f$ turns the line $\ell$ into a "number line"

The number line is the familiar tool from middle-school math: a line with real-number coordinates marked on it, where the distance between two points equals $|x_A - x_B|$. That is exactly what $f$ does:

$f : \ell \;\longleftrightarrow\; \mathbb{R}$

Stick a label written with a real number on every point of the line. Once that is done, the abstract "geometric line $\ell$" becomes a number line.

2. The line is one-dimensional — you can only "slide along the axis" or "flip"

$\ell$ has no width and no "top or bottom side"; it extends infinitely only in two directions. So the only "geometric moves" available in a one-dimensional space are these two:

• Sliding along the axis (pick any point as the origin)
• Flipping the whole line (choose which side counts as positive)

⚠️ There is no rotation in 1D. "Rotating about an axis" is a concept that exists only in 2D and above; in 1D, "rotating by 180°" is the same as flipping, and any other angle is simply meaningless — this is the geometric root of why only translation and flipping remain as degrees of freedom.

3. Distance-preserving ≠ bijective — both requirements must hold

The literal requirement of the axiom:

$|AB| = |f(A) - f(B)|$

• $|AB|$ is the true geometric distance (what you measure with a ruler)
• $|f(A) - f(B)|$ is the absolute value of the difference of the labels

These two numbers must be equal at all times. Be careful to distinguish two things:

Property	Meaning
Bijection	Every point corresponds to a unique real number — none repeated, none missed
Distance-preserving	The absolute difference of the labels = the true distance

$y = x/2$, $y = x^3$, $y = \tan x$ are all perfect bijections $\mathbb{R} \to \mathbb{R}$, but none of them preserve distance. The axiom requires bijection and distance-preservation simultaneously — neither can be missing. The 1.5× stretching counterexample ✗ shown at 4.3 in the video violates distance-preservation, not bijectivity — don't conflate the two.

Putting all three together: every legal $f$ has the form $\varepsilon x + b$

Stacking the constraints "one-dimensional" and "must preserve distance," the mathematical conclusion is: fix any one legal $f_0$; then every legal $f$ on $\ell$ has the form

$f(x) = \varepsilon \cdot f_0(x) + b, \qquad \varepsilon \in \{+1, -1\},\; b \in \mathbb{R}$

Degree of freedom	Type	Geometric meaning	Video chapter
$b \in \mathbb{R}$	1-dim continuous	Translation (choose where zero is)	4.1
$\varepsilon \in \{+1, -1\}$	1-bit discrete	Flip (choose the positive direction)	4.2

This is the isometry group of the line, $\mathrm{Iso}(\mathbb{R})\cong \mathbb{R} \rtimes \mathbb{Z}_2$ (Wikipedia index, UConn Conrad notes). The two animations 4.1 + 4.2 in the video have already exhausted all distance-preserving bijections of $\mathbb{R}$ — there are no other tricks to add.

In practice: why is "without loss of generality, let $A$ be at the origin" always legal?

Later in proofs you will repeatedly see "WLOG let $A$ fall at the origin" or "WLOG suppose $f(B) > 0$" — this is spending the two degrees of freedom of $f$:

• Use $b$ to move $A$ to $0$
• Use $\varepsilon$ to put $B$ on the positive side

The two degrees of freedom are just enough to normalize any pair of distinct points $A, B$ to $f(A)=0,\, f(B)>0$. There is nothing left over; spending any more would break distance-preservation. This is where the legitimacy of WLOG ("without loss of generality") comes from.

In one sentence: $f$ turns $\ell$ into a number line; in 1D only two distance-preserving moves are allowed — translation and flipping — and scaling and distortion are excluded by the axiom. That is why "distance = coordinate difference" can serve as the rock-solid foundation of every geometric argument in middle school.

Immediate consequences

Three things can be read directly off the axiom, with no proof needed:

• Non-negativity: $|AB| = |f(A) - f(B)| \ge 0$.
• Symmetry: $|AB| = |BA|$, since $|x-y| = |y-x|$.
• Zero distance means coincidence: $|AB| = 0 \iff f(A) = f(B) \iff A = B$ (because $f$ is a bijection).

These three facts are the basis of every subsequent distance argument.

Compared with the Elements / Hilbert

Euclid's Elements has no notion of "real numbers." Euclid talks about length through the common notions ("equals added to equals are equal") and repeated "compass-and-straightedge construction of equal segments," but cannot directly say "$|AB| = 3.7$." The result is that any slightly more complicated inequality or limiting argument cannot be expressed cleanly.

Hilbert (1899) patched the gaps in Euclid by purely geometric means — but at the cost of splitting "order," "congruence," "Archimedes," and "completeness" into five groups of axioms (about 20 axioms in all). Birkhoff's Ruler axiom handles four things in a single sentence: the continuity of points on a line, order, additivity of distance, and the Archimedean property. The cost is treating the real numbers as an already-constructed object — which is a worthwhile trade for middle-school geometry, and is also honest with students.

What it unlocks

Several upcoming theorems depend, directly or indirectly, on this axiom:

• Existence and uniqueness of midpoints: given $A$, $B$, there exists a unique $M$ with $|AM| = |MB| = \frac{1}{2}|AB|$.
• Triangle inequality (collinear case): $|AC| \le |AB| + |BC|$, with equality when the three points are collinear.
• Segment arithmetic: segment lengths, as real numbers, can be manipulated algebraically.
• Order: "$B$ lies between $A$ and $C$" on a line is defined as $f(A) < f(B) < f(C)$ or the reverse.

The next axiom, Axiom III · Protractor, will transplant the same structure to angles; afterwards, $\textbf{IV · SAS similarity}$ stitches the two metrics — distance and angle — together, which is enough to prove every geometric theorem in middle school.